| 1. |
- Granström, Karl, 1981, et al.
(författare)
-
Likelihood-Based Data Association for Extended Object Tracking Using Sampling Methods
- 2018
-
Ingår i: IEEE Transactions on Intelligent Vehicles. - 2379-8858. ; 3:1, s. 30-45
-
Tidskriftsartikel (refereegranskat)abstract
- Environment perception is a key enabling technology in autonomous vehicles, and multiple object tracking is an important part of this. The use of high resolution sensors, such as automotive radar and lidar, leads to the extended object tracking problem, with multiple detections per tracked object. For computationally feasible multiple extended object tracking, the data association problem must be handled. Previous work has relied on a two-step approach, using clustering algorithms, together with assignment algorithms, to achieve this. In this paper, we show that it is possible to handle the data association in a single step that works directly on the desired likelihood function. Single step data association is beneficial, because it enables better use of the measurement model and the predicted multiobject density. For single step data association, we use algorithms based on stochastic sampling, and integrate them into a Poisson Multi-Bernoulli Mixture filter. In a simulation study, and in an experiment with Velodyne data acquired in an urban environment, four sampling algorithms are compared to clustering and assignment. The results from the simulations and the experiment show that single-step likelihood-based data association achieves better performance than two-step clustering and assignment data association does.
|
|
| 2. |
- Patel, Raj-Haresh, 1991, et al.
(författare)
-
Buffer-Aided Model Predictive Controller to Mitigate Model Mismatches and Localization Errors
- 2018
-
Ingår i: IEEE Transactions on Intelligent Vehicles. - 2379-8858. ; 3:4, s. 501-510
-
Tidskriftsartikel (refereegranskat)abstract
- Any vehicle needs to be aware of its localization, destination, and neighboring vehicles' state information for collision free navigation. A centralized controller computes controls for cooperative adaptive cruise control (CACC) vehicles based on the assumed behavior of manually driven vehicles (MDVs) in a mixed vehicle scenario. The assumed behavior of the MDVs may be different from the actual behavior, which gives rise to a model mismatch. The use of erroneous localization information can generate erroneous controls. The presence of a model mismatch and the use of erroneous controls could potentially result into collisions. A controller robust to issues such as localization errors and model mismatches is thus required. This paper proposes a robust model predictive controller, which accounts for localization errors and mitigates model mismatches. Future control values computed by the centralized controller are shared with CACC vehicles and are stored in a buffer. Due to large localization errors or model mismatches when control computations are infeasible, control values from the buffer are used. Simulation results show that the proposed robust controller with buffer can avoid almost the same number of collisions in a scenario impacted by localization errors as that in a scenario with no localization errors despite model mismatch.
|
|
